No. of Recommendations: 2
I am currrently investigating MSS with moire or less simple high and low components to get a pattern. While doing some of these endless sessions at Jamies engine I found this switch:

http://gritton.org/ws/switch/?86990101djOVbwbs113l3USTpq15Up1c0

I can't remember any screen for which Jamies engine is reporting such a high sharpe ratio. Is there anything wrong? CAGR is not so impressive, "only" 58%, GSD is very low at 20. Some screen for those who are more risk orientated, than CAGR hunting?

Attention the screen is using more stocks on the high side.

YTD results reduces the GSD another point.

Volker
Print the post Back To Top
No. of Recommendations: 4
http://gritton.org/ws/switch/?86990101djOVbwbs113l3USTpq15Up1c0

I can't remember any screen for which Jamies engine is reporting such a high sharpe ratio [1.98]. Is there anything wrong? CAGR is not so impressive, "only" 58%, GSD is very low at 20. Some screen for those who are more risk orientated, than CAGR hunting?


When I looked at your screen, I noticed that your "low screen" actually underperforms the index during its share of the method. I replaced it with RS4 2-4 (and changed the PEG13 to 2-4 as well):

http://gritton.org/ws/switch/?v86000101djSTrm24USTpq24Up1c0

It increases both CAGR (to 66%) and the Sharpe ratio (to 2.24).
Print the post Back To Top
No. of Recommendations: 0
Randy,

with all reservations to dropping #1 stocks, the simplicity of this screen and its results are really beautiful.

And YTD 00 without the August numbers = 85%!

This one deserves a place in the MSS hall of fame, unless anyone finds something more beautiful.

Volker
Print the post Back To Top
No. of Recommendations: 0
And YTD 00 without the August numbers = 85%!

Volker,

I assume that was a typo, the YTD 00 return is 35%. But I agree, the screen is beautiful in its simplicity.

Print the post Back To Top
No. of Recommendations: 3
Volki,

Check this out:http://gritton.org/ws/switch/?v86000101djSTrm14USTpq24Up1c0

Is this datamining?
Print the post Back To Top
No. of Recommendations: 0
I assume that was a typo, the YTD 00 return is 35%.

No typo please check

http://gritton.org/ws/switch/?86000101djSTrm24USTpq24Up1c0

For the YTD result is 85%

Volker
Print the post Back To Top
No. of Recommendations: 1
Check this out:http://gritton.org/ws/switch/?v86000101djSTrm14USTpq24Up1c0

Is this datamining?


Who knows for sure? At leastz it lloks good to me. Maybe less datamining because it is only one first slot which is dropped. But as time goes by, take a look at this one, and don't tell it to your friends :-)

http://gritton.org/ws/switch/?v86000101djSTrm13USTpq24Up1c0

CAGR 86-2000
76%
GSD 86-2000
20%
Sharpe Ratio 2.21
CAGR/GSD
3.8

Well that is a real gem. The higher GSD happens only on the upside, so don't worry about these two points. These switches are real KISS.

Volker


YTD 2000
102%

I do not hope Jamies backtesting engine is broken.
Print the post Back To Top
No. of Recommendations: 2
YTD 2000
102%


Have to quote myself. The August picks were ESRX MENT SEIC. This port is up since 07/31/200 (closing price) another 10,6% until friday. If there doesn't happen weird things on monday this MSS will be up to approx. 112% YTD.

This is really good.

Volker
Print the post Back To Top
No. of Recommendations: 0
Volker said:

This one deserves a place in the MSS hall of fame, unless anyone finds something more beautiful.

Not as beautifully simple but combining two switch screens, this one and the other famous switch screen look at this one:
http://gritton.org/ws/blend/?vi10000t860001SW01djSTrm24USTpq24Up1c0U50nSW01djSSbwrqlbrs5m5l3UOVpqpw314l3Up1c0.9U50n

CGAR:90
GSD:22
Sharpe 2,10

This one might be a bit more datamined.

FSC
Print the post Back To Top
No. of Recommendations: 1
FSC,

datamine? IMHO it shows the power of the concept of blending. The latest blend would look like;

http://gritton.org/ws/blend/?vi10000t860001SW01djSTrm13USTpq24Up1c0U50nSW01djSSbwrqlbrs5m5l3UOVpqpw314l3Up1c0.9U50n

CAGR 94
GSD 22
Sharpe 2.08

Volker
Print the post Back To Top
No. of Recommendations: 3
http://gritton.org/ws/sos/?v86000101SWdjSTrm13USTpq24Up1c0USWdjSSbwrqlbrs5m5l3UOVpqpw314l3Up1c0.9U10m10l6

CAGR 92
GSD 22
Sharpe 2.18

This SoS prevents doubling up on stocks which appear on both switches. This should reduce the LMS risk a bit.

Jamie,

can you see how are playing with your engines?

Volker
Print the post Back To Top
No. of Recommendations: 0
Volker,

Sorry about the typo remark, I thought you were referring to your original Plow-based screen rather than Randy's RS4-based screen.

Anyway, great work on the MSS screens! Keep it up!
Print the post Back To Top
No. of Recommendations: 0
When is it better to search for a great Sharpe Ratio
vs. a great CAGR and a good Sharpe Ratio?



http://gritton.org/ws/sos/?v86000101OOrqrss12rqpss23pqpss23pq22USWdjSSbwrqlbrs5m5l2UOVpqpw314l2Up1c.9U10m10l5
Print the post Back To Top
No. of Recommendations: 5

Not as beautifully simple but combining two switch screens, this one and the other famous switch screen look at this one:
http://gritton.org/ws/blend/?vi10000t860001SW01djSTrm24USTpq24Up1c0U50nSW01djSSbwrqlbrs5m5l3UOVpqpw314l3Up1c0.9U50n

CGAR:90
GSD:22
Sharpe 2,10

This one might be a bit more datamined.


It may be datamined but it does hold up to the multiple start test, i.e. calculate every annual return in the 175 monthly returns. There are 164 annual periods:

CAGR of 164 periods = 92.8%
GSD of the same = 25.6%

Conclusion - the great results presented are not explained by a lucky choice of the 14 January returns.

Kevin L
Print the post Back To Top
No. of Recommendations: 0
Just for fun in response to #78303:

http://gritton.org/ws/blend/?vi10000t860001SW01djSTrm13USTpq24Up1c0U50nSW01djSSbwrqlbrs5m5l3UOVpqpw325l3Up1c0U50n

The Sharp jumped up to 2.53. Dropped the CAGR by 11%, to 80, but decreased GSD by 22%, to 17.

It's so juicy, I'm almost thinking about using this one myself. Tweet, tweet, tweet, gasp, choke, arghh. But then again, who knows??

:^)

Curt
Print the post Back To Top
No. of Recommendations: 0
Just for fun in response to #78303:

http://gritton.org/ws/blend/?vi10000t860001SW01djSTrm13USTpq24Up1c0U50nSW01djSSbwrqlbrs5m5l3UOVpqpw325l3Up1c0U50n

The Sharp jumped up to 2.53. Dropped the CAGR by 11%, to 80, but decreased GSD by 22%, to 17.

It's so juicy, I'm almost thinking about using this one myself. Tweet, tweet, tweet, gasp, choke, arghh. But then again, who knows??


This canary cage is quite amazing. The lowest annual return during the backtest period was 45%. CAGR-3GSD is +12% !

Elan
Print the post Back To Top
No. of Recommendations: 1
Well, I ran GSD(M) analyses on this stunning thread.

It turns out that all of the 17-23% GSDs become ~35% when all months are taken into account. (I don't know how to reconcile this with klouche's result of ~26% -- I presume it's the difference in methods used).

The best performance is by Volki's blend in #78271, with CAGR/GSD = 94/35. His SOS of the same is barely worse, 92/34, but has the comfort of not doubling up (or worse -- several periods totally overlapped the blend, with a total of only 3 stocks).

I still like the complex half of the blend by itself, yielding 116/41.
Replacing PEGo with SOS-G (kind of adding RS52 to PEGo) gives 123/44.
Finally, StephenFool's SOS for the PEGo side, which mixes in RSCAP and PlowRSW, gives 129/44. Tweet, tweet, but holy cow!!

-Jim
Print the post Back To Top
No. of Recommendations: 2
Combine OCCM + msg.78262 and you get
sharpe of 2.37 (the highest yet??!!)
CAGR: 86 GSD: 19 YTD 2000: 74

http://gritton.org/ws/sos/?v86000101OOrqrss12rqpss23pqpss23pq22USWdjSTrm13USTpq24Up1c0U10m10l4
Print the post Back To Top
No. of Recommendations: 0
Author: Volki     Date: 8/27/00 11:31 AM    Number: 78206
<<<..........
Attention the screen is using more stocks on the high side. ......................>>

If you set the pq limit to l3 you get:
3 stocks high and low

Switch on DJIA last period, 0% cutoff, high: (PEG13 1-3), low: (PlowRSW/Plow overlap 1-3:3, pad plowrsw)
http://gritton.org/ws/switch/switch.csv?v86990101djOVbwbs113l3USTpq13Up1c0

Cagr 67%
GSD 30.3
Sharpe Ratio 1.5

Best!!!
Print the post Back To Top
No. of Recommendations: 0
Author: centralbev     Date: 8/27/00 3:12 PM    Number: 78257


I like to have the portfolis to hold the same number of stocks.

Try this!

Switch on DJIA last period, 0% cutoff, high: (PEG13 2-4), low: (RS4 1-3)
http://gritton.org/ws/switch/switch.csv?v86990101djSTrm13USTpq24Up1c0

CAGR 73.9
GSD 20.2
Sharpe 2.12
Print the post Back To Top
No. of Recommendations: 0
I have been following the various discussions of MSS with great interest but I am left with a question: if one actually implements such a screen using real money, then what dates are used to determine the change in the DJIA or other indicator for switching? Off the top of my head, I would be inclined to use something like the closing price on the previous rebalance day and the openning price on the current rebalance day. Does that make sense? Do others use something different?

Apologies if this sounds naive and/or has been addressed in a post that I missed.
Print the post Back To Top
No. of Recommendations: 1
tpto said:
When is it better to search for a great Sharpe Ratio
vs. a great CAGR and a good Sharpe Ratio?

http://gritton.org/ws/sos/?v86000101OOrqrss12rqpss23pqpss23pq22USWdjSSbwrqlbrs5m5l2UOVpqpw314l2Up1c.9U10m10l5


It's easier to get very high CGAR's when you only have 2-3 stocks, go to 6 to 20 stocks and try...

see this:

http://gritton.org/ws/blend/?vi10000t899901SW01djSTrm13USTpq13Up1c0U34nST01ke14U10nSW01djSSbwrqlbrs5m5l3UOVpqpw314l3Up1c0.9U34nST12ss15U14nST01hp12U8n

This is a combination of screen switch and a touch of Kuperman foundation blend.
I've included also HighPrice, so my start date begins in 1989.

CGAR:94
GSD 31
Sharpe:1.67

If you run it from 1986 the figures go a bit down, but keep in mind that the money corresponding to highprice is not invested.

CGAR 85
GSD 29
sharpe: 1.59

And the performance of the screen's YTD: 53


This is a diversified screen with the majority of it's months between 12 and 16 stocks.

Francisco







Print the post Back To Top
No. of Recommendations: 1
OK. This thread is excellent. But the question is, how the heck does one calculate what stocks to buy with one of these complex MSS's. To the best of my knowledge these types of screens are shown in any of the ledgers - etc.

To invest in them you would need a pretty heavy duty set of rule laid out...

Cory
Print the post Back To Top
No. of Recommendations: 0
Cory said:
OK. This thread is excellent. But the question is, how the heck does one calculate what stocks to buy with one of these complex MSS's. To the best of my knowledge these types of screens are shown in any of the ledgers - etc.

Jus go to Gritton backtest every last friday/saturday of the month, run the blend with the start date set at the month of your port start date and with the end year on the current year. then pick the current month ranks (you have to check "list picked stocks" in Jamie's engine)

Francisco
Print the post Back To Top
No. of Recommendations: 0
"But the question is, how the heck does one calculate what stocks to buy with one of these complex MSS's."

I agree that some of these things are difficult to program in Excel (with my limited skills anyway). However, if the screens depend on selecting stocks from Jackcade's short list, then a program to select the stocks can be implemented in a data processing package like SAS. I have a few routines set up to select stocks for various overlaps and SOS approaches. Would be happy to share same. Also wouldn't mind having someone else check the code.
Print the post Back To Top
No. of Recommendations: 0
This is a diversified screen with the majority of it's months between 12 and 16 stocks.

For now I'm ok with 3 to 5 stocks in this part of my
portfolio since it's only about 1/4 of my net worth.
(I include house equity, cash, company stock options,
401Ks stuck in index funds etc.)

I have been using the Purple Egg SoS for the
last five months for reasons explained elsewhere. It
has returned ~15% in five months. Since it is my
favorite SoS I decided to blend it with one of the
recently touted MSS screens.

I report the 86-98 results because the 635% gain in
1999 plays hob with the Sharpe Ratio (unfairly I
think).

CAGR: 106
GSD: 23
Ratio: 2.28

The minimum gain ('92) is 51%, 7 of 14 years greater
than 100% (not including '99). Almost half the months
pick 4 stocks, the other half are equally divided
between 3 and 5 stocks. Maximum drawdowns etc. look
pretty good too.


http://www.gritton.org/ws/blend/?v869801OV01SSSTpq26USTra14USTpw15USTbw15U10m7l10USSSTpq26USTpw15USTrq14USTra14USTbw15U10m7l10U013l3U2nSW01djSSbwrqlbrs5m5l3UOVpqpw314l3Up1c1U2n

Peter
Print the post Back To Top
No. of Recommendations: 3
Jus go to Gritton backtest every last friday/saturday of the month, run the blend with the start date set at the month of your port start date and with the end year on the current year...

But for MSS, be careful: I only get the index values in after closing Friday. Currently, it'll give bogus picks if you look too early (I need to fix this as was suggesteed recently).

- Jamie
Print the post Back To Top
No. of Recommendations: 0
I have been following the various discussions of MSS with great interest but I am left with a question: if one actually implements such a screen using real money, then what dates are used to determine the change in the DJIA or other indicator for switching?

One possibility is what I actually used in the backtests: compare closing prices on the Friday the rankings come out vs the Friday close from the previous month, and buy the stocks on Monday.

For my own portfolio, I do the same off one day: I compared Thurday closes and bought on Friday. Last month that made a difference: the OEX dropped sharply on 28 Jul, and I ended up with the "high" screen while the Friday-close would have mandated the "low" screen. For that matter, the OEX's current value at the time I picked the stocks was already pointing low. So expect your results to vary.

- Jamie
Print the post Back To Top
No. of Recommendations: 1
OK. This thread is excellent. But the question is, how the heck does one calculate what stocks to buy with one of these complex MSS's. To the best of my knowledge these types of screens are shown in any of the ledgers - etc.

To invest in them you would need a pretty heavy duty set of rule laid out...


If you are using SOS's, this SOS calculator is very useful. You can construct an SOS and bookmark the URL. All you need do is click on your bookmark each rebalance day after Brian Finney has updated his site. It uses the latest data from Brian Finney's site.

http://www.halcyon.com/blm/ws/saucer/

Stephen
Print the post Back To Top
No. of Recommendations: 0
Just for fun in response to #78303:

http://gritton.org/ws/blend/?vi10000t860001SW01djSTrm13USTpq24Up1c0U50nSW01djSSbwrqlbrs5m5l3UOVpqpw325l3Up1c0U50n

The Sharp jumped up to 2.53. Dropped the CAGR by 11%, to 80, but decreased GSD by 22%, to 17.

It's so juicy, I'm almost thinking about using this one myself. Tweet, tweet, tweet, gasp, choke, arghh. But then again, who knows??

This canary cage is quite amazing. The lowest annual return during the backtest period was 45%. CAGR-3GSD is +12% !

Elan


I wonder how this one has held up...can someone convert the string for backtest.org?

TIA. Happy Thanksgiving MI. We have much to be thankful for.

Tails
Print the post Back To Top